In collaboration with E. Green, who has developed a detailed YAC-based, STS-content map of human chromosome 7, we propose to sequence blocks of genomic DNA on the following scale; year 1, 2 Mbp; year 2, 6 Mbp; year 3, 20 Mbp. Initial efforts will be directed at the region of the human homolog of the mouse ob gene in 7q31.3, where a YAC contig of more than 3 Mbp is available. Individual sequenced segments will be more then 1 Mbp in all cases. We will aim for complete, contiguous sequence. We estimate error rates of significantly les s than one error per thousand base pairs, with a high proportion of the errors having little likelihood of affecting biological interpretations of the data (e.g., miscounts of the number of nucleotides in long homopolymer runs). We expect on the order of a few short discontinuities per megabasepair or less. All sequence tracts will be aligned and oriented relative to the chromosome map, and estimates of gap sizes will be made. Sequence-ready cosmid clones and maps will be produced by multiple-complete-digest mapping. The cosmids will be derived by a mixed strategy of probing a chromosome-specific cosmid library with labeled YAC DNA and subcloning the YACs. This strategy combines protection against propagation of YAC-cloning artifacts with the ability to obtain deep, targeted coverage, while freely experimenting with varied cloning systems. Sequencing will be by shotgun sampling of the cosmids followed by universal-primer, four-color-fluorescence, dideoxy sequencing of the subclones. This project will represent the first intensive field test both of multiple-complete-digest image-analysis and map-assembly software, as well as base-calling and sequence-assembly software, all of which have been developed and carried through initial testing in our Genome Center. Initial results have been highly promising, and we expect these software packages to decrease greatly the labor-intensiveness of high-resolution mapping and genomic sequencing while still yielding a high quality product. Practical refinement of these tools is a major goal of the proposed Project, and we believe such refinement can only proceed optimally with the combination of intimate involvement of the developers, strong software engineering, and an ambitious data-production effort that we propose.